GeneSelectR (version 1.0.1)

plot_upset: Plot Feature Overlaps Using UpSet Plots

Description

This function produces separate UpSet plots for inbuilt feature importances and permutation importances, allowing you to visualize the overlap of feature lists. Optionally, you can include custom lists.

Usage

plot_upset(pipeline_results, custom_lists = NULL)

Value

A named list containing two UpSet plots:

• @field inbuilt_importance: An UpSet plot visualizing overlaps of inbuilt feature importances.

• @field permutation_importance: An UpSet plot (if permutation importance is available) visualizing overlaps of permutation importances. Each plot provides an interactive way to explore the intersections and unique elements of the feature lists.

Arguments

pipeline_results

A PipelineResults object containing the fitted pipelines, cross-validation results, selected features, mean performance, and mean feature importances.

custom_lists

An optional named list of character vectors. Each character vector should contain feature names. The names of the list will be used as names in the UpSet plots.

Examples

# \donttest{
# Mock data for PipelineResults
pipeline_results <- new("PipelineResults",
                        inbuilt_feature_importance = list(
                          Method1 = data.frame(feature = c("gene1", "gene2", "gene3")),
                          Method2 = data.frame(feature = c("gene2", "gene4"))),
                        permutation_importance = list(
                          Method1 = data.frame(feature = c("gene1", "gene5")),
                          Method2 = data.frame(feature = c("gene3", "gene6"))))

# Mock custom lists
custom_lists <- list("custom1" = c("gene1", "gene2"), "custom2" = c("gene3", "gene4"))

# Generate UpSet plots
result <- plot_upset(pipeline_results, custom_lists)
print(result$inbuilt_importance)
print(result$permutation_importance)
# }